Typically, computer vision applications (e.g. three-dimensional depth sensing applications, etc.) require two cameras to operate. Specifically, such cameras include at least one dedicated infrared camera for generating a three-dimensional depth map, and at least one camera for capturing a color image. Further, devices such as mobile phones now more typically include at least one camera on both a front and back face of the device.
Thus, in order for such devices to support the aforementioned applications using both the front and back face camera(s), there is a requirement for a total of four (4) cameras per device (i.e. two (2) on the front face and two (2) on the back face). However, with a continuing desire to reduce a size of such devices, there is not enough space to fit such four cameras per device. Further, previous attempts to incorporate both infrared and color pixels into a single sensor fail to support other techniques (e.g. binning, etc.) that boost signal-to-noise ratio (SNR) performance.